Because it is executed by CPU.

So it really depends on the hardware instruction set.

Generally speaking, int type addition needs to be done once.

Multiplication is about 5 times that of addition, depending on the kernel instruction implementation.

The fastest may be 2 times and the slowest may be 40 times.

Floating-point operation mainly depends on whether there is a special hard floating-point operation unit. If there is, it is similar to addition, a little slower.

If not, it will be much slower.